Single Fraction vs. Fractionated Preoperative Radiosurgery for Resected Brain Metastases: A PROPS-BM Multicenter Cohort Study

Abstract

<h3>Purpose/Objective(s)</h3> Preoperative (preop) stereotactic radiosurgery (SRS) is a feasible alternative to postoperative (postop) SRS with potential benefits in adverse radiation effect (ARE) and meningeal disease compared to postop SRS. Most reported studies of preop SRS utilized single fraction (fx) SRS. The goal of this study was to compare outcomes and toxicity of single fraction preop SRS with fractionated (3-5 fx) preop stereotactic radiotherapy (FSRT). <h3>Materials/Methods</h3> Patients (pts) with brain metastases (BM) from solid cancers, of which at least 1 lesion was treated with preop SRS or FSRT and underwent planned resection were included from 6 institutions. SRS to synchronous intact BM was allowed. Exclusion criteria included classically radiosensitive or non-solid cancers and prior or planned whole brain radiotherapy. RT dose, fractionation, and interval between preop RT and surgery was per individual institutional protocol. Intracranial outcomes were estimated using cumulative incidence with competing risk of death. Propensity score matched (PSM) analyses were performed. <h3>Results</h3> The cohort included 378 pts with 389 preop RT treated index lesions. Most pts had non-small cell lung (47.9%), breast (16.1%), or melanoma (11.4%) cancer. The majority of pts (61.1%) had a single BM and 95.4% underwent gross total resection (GTR). SRS was used for 328 (84.3%) lesions with FSRT used for 61 (15.7%). Most FSRT (93.4%) was 3 fx, with the remainder being 5 fx. Median prescribed dose for FSRT and SRS was 24 Gy and 15 Gy, respectively. There were baseline imbalances between cohorts. The FSRT cohort had significantly more use of planning target volume (PTV) margin expansion > 1 mm (64% vs. 5.2%), more pts with multiple BM (54% vs. 38%), and larger median gross tumor volume (GTV, 12.3 cc vs. 9.7 cc). In univariate analysis, preop FSRT was associated with significantly lower rate of cavity local recurrence (LR) than SRS (1-year: 2.8% vs. 14.1%, p=0.02). PSM yielded 57 matched pairs. There were no significant residual imbalances in the PSM cohorts. Median imaging FU period was 8 months for both groups. FSRT was associated with significantly lower cavity LR compared with SRS (1-year: 3% vs. 17.9%, p=0.02). There were no differences in ≥ grade 2 ARE (7.4% vs. 4.9%, p=0.34), meningeal disease (2% vs. 4%, p=0.81), or distant brain failure (21.8% vs. 37.4%, p=0.25). No grade 3 - 5 ARE events occurred. Overall survival (OS) was significantly higher with FSRT in PSM analysis as well (1-year: 71.2% vs. 52.1%, p=0.04). <h3>Conclusion</h3> Preop FSRT (3-5 fx) is associated with significantly reduced risk of cavity LR and improved OS compared with preop single fx SRS. Both preop FSRT and SRS show notably low rates of ARE and meningeal disease. Three fx preop FSRT may be a preferred option for neoadjuvant RT of larger or higher risk lesions. Additional confirmatory studies are needed. A randomized trial of preop SRS versus postop SRS is currently being designed (NRG BN012).